STATISTICAL COMPARISONS. 471 



the school year includes 10 months, the average is 7"l7 months ; 

 in Rhode Island, 9 months 11 days ; and in New Jersey, 9 

 months 10 days. It is manifest, therefore, that the figures 

 representing the regularity of attendance require material 

 correction and reduction before they can be properly compared 

 with the statistics of European countries in which schools are, 

 as a rule, kept open during nearly the whole of every year." 



From this it is quite clear that one has the greatest difficulty 

 in discussing such a question as the education of a people. You 

 can hardly get to know to what extent children of school age 

 are attending schools of some kind. Thei'e are other difficulties 

 behind, as the rejDort from which I have quoted shows, 

 such as the diffi^rence of surroundings in which children find 

 themselves when they leave school, the United States, from 

 the general vigour and energy of the whole population, being 

 much more favoui'able to the development of general in- 

 telligence and mental cultivation among its people than countries 

 which may be more fortunate as regards primary school educa- 

 tion. There is also such a difficulty as the kind and character 

 of secondary education, and the extent to which it is diff'used. 

 Simple at first sight as the problem seems, then, there is 

 nothing more difficult than to compare some countries with 

 each other as regards the degree of their education. 



The second subject I have named in this connection is crime, 

 and in thinking of it I confess I have had in mind certain com- 

 parisons which have been made in England by visitors returned 

 from Australia to the disadvantage of Australia. There is twice 

 the crime in Australian colonies per head of population, we have 

 been told, that there is in England. But, as we all know who have 

 to handle statistics, there are few statistics so difficult to handle 

 as those of crime. A distinction has to be made between mere 

 police and adrainisti'ative offences, which vary largely according 

 to the things which Legislatures in their wisdom subject to fine 

 or not, and the more serious oftences, such as robbery and 

 murder, which are what we think of when we talk of crime. 

 But in hardly any two countries that I know of is the distinction. 

 drawn on exactly the same lines. You are almost never quite 

 sure, therefore, what you are doing, unless you are specially 

 careful, when you compare two countries as regards crime. 

 Further, even if the distinctions were much the same, another 

 difference is made by the police. You may have fewer trials 

 and convictions in one countiy than in another, simply because 

 the police for various reasons is less efficient, not because there 

 is less crime. When comparisons, therefore, are made between 

 the criminal statistics of two countries without attention to 

 vital considerations like these to show that the subject has been 

 really studied, it is safe to dismiss them without further thought. 



But admitting that exact comiDarisons can be made, that 

 statistics of crime in two countries are reduced to common 

 denominators, I should like to point out that the logic of using 

 them as indicative in any way of the general superiority of one 

 population over another may be at fault. So far as can be 

 judged, the so-called crime statistics of a country are not 

 necessarily significant very much of the general quality of a 



